My journey into AI and Machine Learning started with one of my college teachers. She introduced me to AI/ML and made me curious about what could actually be achieved with it. She no longer teaches at our college, but the interest she created in me stayed.
At first, I thought AI/ML was simply something cool and interesting. As I started studying it seriously, I realized that it was much deeper than I had imagined. I became particularly interested in Deep Learning because of how neural networks can learn complex patterns from data and solve problems that are difficult to approach with traditional programming.
The more I learned, the more I realized how much is possible with AI/ML. This project became an opportunity to take what I had learned and apply it to a meaningful problem: brain tumor segmentation from MRI scans.
Building NeuroSeg AI allowed me to explore deep learning beyond theory—from understanding MRI data and preprocessing, to training a segmentation model, evaluating its performance, deploying it as an interactive application, and adding Grad-CAM explainability to understand where the model is focusing.
The project also taught me that building an AI system is not just about achieving a high metric. It involves understanding the data, handling failures, testing on unseen examples, interpreting predictions, and being honest about the limitations of the model.
What started as something I thought was simply “cool” has become something I genuinely want to explore much more deeply. NeuroSeg AI is one step in that journey